27
Li Baowen Big Data Based Maintenance, BDBM--The Main Trend of Smart Maintenance

Big Data Based Maintenance, BDBM--The Main Trend of Smart ...exicon.website/uploads/editor/omaintec2018/presentations/03 - S11-… · Big Data Based Maintenance--BDBM . Self-learning

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

Li Baowen

Big Data Based Maintenance, BDBM--The Main Trend of

Smart Maintenance

Future Development of Smart Maintenance

OMAINTEC 2

Web safety intelligent managing

Intelligent MRO managing

Maintenance Robot

Intelligent failure diagnosing

Intelligent spare part managing

Intelligent strategy generation

Intelligent inspection and monitoring

Intelligent lubrication

Intelligent information processing

Intelligent knowledge managing

Intelligent repair instruction, IETM

Intelligent training 智能维护 总体框架

Framework Smart

Maintenance

What is BDBM

OMAINTEC 3

Input Output

• From PLC or DCS

• From daily inspection

• From historical records

• From Condition

Monitoring

• ……

• Maintenance policy

• Maintenance standard

• Maintenance schedule

• Possible changed spare

parts

• ……

Big Data Based

Maintenance

Multi-channel input,multi-dimension output! An accurate maintenance package should be delivered

Why shall we introduce BDBM?

OMAINTEC 4

The problem of preventive maintenance:

Time based maintenance:According to current statistics, Only 20%

preventive maintenance is effective, 80% is invalid, even induce new

failures;

Condition based maintenance:Could avoid the defect of TBM,the

input-output ratio should be evaluated;

The current status:

Although some data from the technological control system could reflect

the degradation of the plant, but not being used at all;

Main problems of current situation:

Much useful data is not utilized, much relative datum are not integrated.

The category suitable BDBM

OMAINTEC 5

Distribution of Maintenance Modes

All plant

Obvious exhausting interval Plant or part of it exists

obvious exhausting interval, the regulation is handled by us.

The other parts will run

normally within this period.

The consequence is not serious

Small economic lost Zero accident

No environmental damage No quality damage

No health affect No obvious chain damage

No measure to monitor the

failure

Suited to BDBM

Yes? Yes?

Breakdown maintenance is the most economical

strategy

Time based maintenance is the

best strategy

Yes Yes No No

Introduce the Risk management

OMAINTEC 6

Definition of Risk

Risk=Probability × Consequence=P ×C

Failure probability is complementary to reliability;

P=1-r

Where, C means consequence,C∈[0,1]

r means reliability,r ∈[0,1]

P is the probability of failure,P ∈[0,1]

Risk Evaluation and Elimination

OMAINTEC 7

Focus to Evaluation and Elimination:

Event Identification

Define the basic event

Event Frequency Event Tree

Probability Matrix

Risk Evaluation and Elimination

Consequence Tree

Event Consequence

Accident Event

Relationship of Risk via Maintenance strategy

OMAINTEC 8

0.07 0.14 0.28 0.42 0.49

0.05 0.16

0.04 0.28

0.12

0.01 0.04 0.05 0.06

0.7 disaster >500 million

>100 death

0.6 serious >100 million

>10 death

0.5 key >10 million >1death

0.4 big > million =1 death

0.3 general >100 k Labor hour lost

0.2 Small >10 k Medical treatment

0.1 Very small <10 k hurt

image economic safety

0.1 0.2 0.3 0.4 0.5 0.6 0.7

Impossible at all

impossible

few Low possible

Very possible

often

1/100000

1/10000

1/1000Y

1/100Y

1/10Y

1/Y 10/Y

Generator set

Part 1 *

Part 2 *

*

Part n *

Turbine set Part 1 *

Part n * *

Boiler

Part 1 *

*

Part n *

DCS Daily inspection

Condition monitoring

Historical record

Other

Data resource

strategy

BDBM x x x x x

CBM x x

TBM x x x

PIT STOP

x x x x x

OppT x x x x - - -

ProA - - - x x x

BrkD x x x x - - -

Risk matrix Consequence matrix

Probability matrix

Data resource matrix

Maintenance stratety

matrix

OMAINTEC 9

Some concept of Risk Probability of Series System:

P=1-∏ri

Where,ri represents the reliability of the ith sub-system. The

reliability will be lower while more and more sub-systems

are series connected 。

Probability of Parallel System:

P=∏(1-ri)

Where,ri represents the ith parallel sub-system, The

reliability will be higher while more and more sub-systems

are parallel connected 。

OMAINTEC 10

Color Management of the Risk

Risk Evaluation and Color Management

Grade of Risk Value of Risk Color

Intolerant 0.5-1

Serious 0.2-0.5

General 0-0.2

OMAINTEC 11

Risk Map of Turbine

Assuming case: Risk of Turbine

High

Serious

General

OMAINTEC 12

Risk Map of Generator

Assuming case: Risk of Generator

High

Serious

General

OMAINTEC 13

Risk Map of Generator

Time

Peak period

Ordinary time

High

Serious

General

Assuming case: Risk of Generator in deferent time

OMAINTEC 14

Risk Map of Escalator in Metro

High

Serious

General

Assuming case: Risk of Escalator Of Metro

OMAINTEC 15

Risk Map of Escalator in Metro

Time

Holidays

Ordinary time

Big event

Rush hour

High

Serious

General

Assuming case: Risk of Escalator Of Metro in different time

OMAINTEC 16

The Significant to set up the Risk Map The Significant:

Let higher leader focus to High Risk Area or Equipment ;

Reinforce the investment to the high risk area or

equipment;

BDBM will first choose the High Risk Area or System to

test;

Risk is a dynamic concept, changing with different

loading time, different age of the equipment. So,

reviewing and drawing a Risk Map every half year is

necessary.

OMAINTEC 17

Big Data Based Maintenance--BDBM Logic Process:

Determine the high risk equipment;

Extract the Feature Value according to different failure of equipment ;

Design the feature value trap threshold X-day, Y-day and Z-day before the

breakdown occurs based on historical failure process ;

Monitoring the feature value continuously or with high frequency;

Minus 50% Feature Values drop in the “trap”, then a light warning start, and

a predictive period is given;

Exact 50% Feature Values drop in the “trap”, then a middle warning start,

and a predictive period is given;

All Feature Values drop in the “trap”, then a strong warning start, and a

predictive period is given;

A maintenance decision is made and a maintenance package is generated.

OMAINTEC 18

Big Data Based Maintenance--BDBM

Feature value extracting

Historical record PLC/DCS Condition monitoring Inspecting

Trap value ranging

Light warning

Middle warning

Strong warning

Maintenance package generation

Monitoring

Failure prediction

N N

Y Y

Failure prediction

What

Adress or equipment to maintain

Where

Maintenance technician or team

Who

Package number and reason

Why

Allocation or repair point

Which

Time and period for maintenance

When

Content of maintenance

Safety Instruction

Safety

OMAINTEC 19

Big Data Based Maintenance--BDBM BDBM--Case Simulation:

From the failure record, we extract 3 feature values :Voltage,

Temperature and Flow, and get the changing trend curve like below:

-30 day -7 day -15 day Failure time

Voltage

Temperature

Flow

From condition monitoring

From DCS control system

From mobile inspection

t

Equipment condition

record

OMAINTEC 20

Big Data Based Maintenance--BDBM

-30 day -7 day -15 day

Tr30-1

Tr30-2

Tr30-3

Tr15-1

Tr15-2

Tr15-3

Tr7-1

Tr7-2

Tr7-3

t Failure time

Voltage

Temperature

Flow

Equipment condition

record

BDBM--Case Simulation:

We design 3 threshold on the curve before 30, 15 and 7 days as traps

to monitor the future status of equipment.

OMAINTEC 21

Big Data Based Maintenance--BDBM BDBM--Case Simulation:

During the process of monitoring, we discover“Flow”get intoTr30-1, then

the light warning is started, and “the failure will happen within 30 days” is

predicted;

While 3 feature value all get into traps, then the strong warning is started,

and “the failure will happen within 7 days”, and the maintenance package is

generated.

Tr30-1

Tr30-2

Tr30-3

Tr15-1

Tr15-2

Tr15-3

Tr7-1

Tr7-2

Tr7-3

-30 day -7 day -15 day t Failure time

Voltage

Temperature

Flow

Equipment condition

record

OMAINTEC 22

Maintenance Package--Output of BDBM

What

Adress or equipment to maintain

Where

Maintenance technician or team

Who

Package number and reason

Why

Allocation or repair point Which

Time and period for maintenance When

Content of maintenance

Safety Instruction Safety

6W2H1C

OMAINTEC 23

Maintenance Package--Output of BDBM Three optional maintenance modes of dynamic maintenance package:

Time

Restoring limit

Time

Time

Safety warning

limit

Restorative maintenance Part replacement maintenance

Upgrading or proactive maintenance

High maintenance frequency with degrading

performance

Maintenance frequency keep constant with periodic restoration of performance

Maintenance frequency reduced with improving

performance

Performance

OMAINTEC 24

Big Data Based Maintenance--BDBM

Self-learning process of BDBM:

Maintenance execution according to the package

Change feature

Leaning again

Replenish new one

Redesign the threshold

Monitoring continuously

N

Y

N

N

Y

Y

Y

N

OMAINTEC 25

Three Transformation of BDBM

3 transformation of BDBM :

1 Orders awaiting Accurate delivery

2 Guaranteed by quantity Guaranteed by speed

3 Reactive action Proactive action

Human, material saving

Breakdown shorten

Accident reduce

OMAINTEC 26

BDBM is the core of Smart Maintenance

Text in here

CBM

Artificial Intelligence

Data Integration

BDBM

OMAINTEC 27

Questions and answers!

Welcome to our webpage: www.tnpm.org

Thanks

Leader of Smart Maintenance

Specialty comes from dedication Excellency hails from creation

Tel:400-104-0028

W e C h a t : s h a r e f o r d 1 2 3

Address:20 Yudalong Build. 41, Lujingxi Rd. Guangzhou, China